No US election campaign in living memory has seen as many invocations of “populism” as this one. Both Donald Trump and Bernie Sanders are labelled as “populists;” the term is regularly used as a synonym for “anti-establishment,” irrespective of any particular political ideas; it is also associated with particular moods and emotions: populists are “angry,” their voters are “frustrated,” or suffer from “resentment.” Similar claims are made about figures in Europe: Marine Le Pen and Geert Wilders are most commonly referred to as populists, as are Syriza in Greece and Podemos in Spain. Yet do all these political actors actually have anything in common? If we hold with Hannah Arendt that political judgment is the capacity to draw proper distinctions, the widespread conflation of right and left when talking about populism should give us pause. Might the popularity of diagnosing all kinds of different phenomena as “populism” possibly be a failure of political judgment? In fact, could contemporary diagnoses of populism turn out to be a second coming of theories of totalitarianism, where the extremes are always bound to touch each other?

A proper understanding of populism has to avoid two common traps: first, a semantic trap. It is often assumed that a plausible concept of populism has to cover all historical instances when political movements called themselves populists. Hence the widespread tendency in the US to take the agrarian People’s Party — the first instance when activists in North America adopted the word “Populist” — as a starting point. It then follows that populism must have something to do with frustration about the major parties and finance capital (many of the slogans of the People’s Party could have been coined by Occupy Wall Street: “ninety and nine in hovels bare, the one in a palace with riches rare”). According to the same logic, though, populism must have something to do with peasants — and it is no accident that at least until the 1970s social scientific definitions of populism included a reference to agrarian interests. In the US, unlike in most of Europe, “populism” retains positive connotations for many on the left, because the late nineteenth-century populists pursued a progressive agenda in economics and politics (fighting for the secret ballot and the direct elections of senators, for instance). However, social and political theorizing about populism cannot just amount to a decal of one historical phenomenon, as the German sociologist Helmut Dubiel once put it, when criticizing the use of “populism” among American scholars. And it cannot mean that a concept has to apply to all actors who have claimed a particular word for themselves; otherwise, no proper account of socialism would be complete without National Socialism.

The second trap is the tendency to conflate the content of what after all is an “-ism” — that is to say: a political belief system — with the psychological states of its supporters. It is correct that in Europe less educated males are the main constituency of what is commonly referred to as populism; it is true that in surveys many voters of anti-establishment parties register their sense that the country as a whole is declining (an assessment that does not necessarily depend on their personal economic situation, as a number of studies have shown); and it is empirically verifiable that many feel elites are “robbing” them of their own country. But all this is like saying that we best understand Social Democracy if we keep re-describing their voters as workers envious of rich people. The profile of supporters of populism obviously matters, but it is patronizing immediately to reduce all they think and say to “resentment” and explain the entire phenomenon as an inarticulate political expression of the supposed “losers of modernization” (this approach was also typical of mid-twentieth century US liberals invested in modernization theory, such as Seymour Martin Lipset, Edward Shils and Victor C. Ferkiss; especially the latter portrayed late-nineteenth century US populism as the precursor of McCarthyism — or even of a distinctly American variety of fascism).

Here is what I think we should do instead to capture populism as a distinct ideology. It is a necessary but not a sufficient condition to be critical of elites in order to qualify as a populist. Otherwise, anyone criticizing the status quo in, for instance, Greece and Italy would by definition be a populist — and, whatever else one thinks about Syriza and Beppe Grillo, it’s hard to deny that their attacks on old elites were often justified. In addition to being anti-elitist, populists are always anti-pluralist. Populists claim that they, and only they, represent the people. As, for instance, Recep Tayyip Erdoğan put it at a party congress, addressing his critics in Turkey:

Other competitors are just part of the immoral, corrupt elite, or so populists say when running for office; when in government, they will not recognize anything like a legitimate opposition. The populist logic also implies that whoever does not really support populist parties might not be part of the proper people to begin with. As Claude Lefort put it, according to this logic, the real people first have to be “extracted” from the people. The proper people, then, is by definition morally pure and unerring in its will (which the populist leader claims just to implement, as with an imperative mandate).

Populists in opposition obviously have to explain why, if their core claim to representativeness is correct, they aren’t in power already. This is where the virtually inevitable appeal to the “silent majority” comes in: if the majority weren’t silent or in some way oppressed by the currently powerful elites, the populists would have long won. When they lose elections, populists then often begin to question the existing political institutions, which are obviously producing the wrong outcome, or even accuse the winners of fraud, as Donald Trump just did. At the very least, they tend to distinguish an empirical and a moral outcome of a ballot. Think of Victor Orbán claiming after losing the 2002 Hungarian elections that “the nation cannot be in opposition,” or think of Andrés Manuel López Obrador, arguing, after his failed bid for the Mexican presidency in 2006, that “the victory of the right is morally impossible” (and declaring himself “the legitimate president of Mexico”).

Populism is the permanent shadow of modern representative politics. As long as we have not moved into what some scholars are anticipating as a “post-representative” era, democracy invites the claim that current elites are not really representative. In fact, as the Yale political theorist Bryan Garsten has argued, representative democracy is designed to multiply claims to representativeness. But, crucially, such claims do not have to be made in an anti-pluralist idiom. Sooner or later they also have to be verified with existing democratic procedures. And that still means, above all, elections. Losing elections will not prevent populists from speaking in the name of “real Americans,” as populists from George Wallace to Donald Trump have done, and to play off a symbolic representation of the authentic people against the empirically verifiable expression of majority will in elections. But then they should also be called out for what they are: not the true disciples of democracy, who want to implement an undistorted popular will, but in fact a danger to democracy.